基于无线传感器网络的产前母猪行为监测系统

发布时间：2018-09-03 20:13

【摘要】：目前大规模集约化母猪养殖业主要是依靠饲养员对产前母猪行为的连续观察,通过直觉和经验从而判断母猪的分娩时间。但是该技术不仅要求大量人力,效率低下,而且往往会出现人为疏忽造成刚分娩的仔猪死亡现象。针对这一问题,设计了一种基于无线传感器网络的产前母猪行为特征监测系统。该系统通过给母猪佩戴加速度传感器采集节点,实时监测母猪的行为,并通过支持向量机分析加速度数据并建立模型,分类识别母猪的站立、躺卧、进食、筑巢等典型行为特征。试验结果表明,系统能够实时连续的记录母猪产前行为特征参数,可准确识别出母猪的四种典型行为,准确率为93.3%。
[Abstract]:At present the intensive sows breeding industry mainly depends on the continuous observation of the behavior of prenatal sows by breeders and judges the delivery time of sows by intuition and experience. But this technique not only requires a lot of manpower and inefficiency, but also leads to the death of piglets. In order to solve this problem, a Prenatal sows behavior monitoring system based on wireless sensor network (WSN) is designed. The system can monitor sows' behavior in real time by using acceleration sensor to collect nodes, and analyze acceleration data by support vector machine (SVM) and establish models to classify and identify sows' standing, lying down, feeding, and so on. Typical behavioral characteristics such as nesting. The results show that the system can record the characteristic parameters of sows' prenatal behavior continuously and accurately identify the four typical behaviors of sows, and the accuracy is 93.33.
【作者单位】：陕西科技大学电气与信息工程学院;
【基金】：陕西省农业科技创新与攻关项目(2015NY028) 西安市未央区科技计划项目(201305) 陕西科技大学博士科研启动基金(BJ13-15)
【分类号】：S828;S818.9

【参考文献】